I've been maintaining a tool that does similar: https://github.com/emberian/cv. High performance parsers of the major harness formats at several versions. Not as much autosyncing. Great results using custom prunings compared to compaction summarization.